US President Donald Trump, after meeting with top executives from the largest technology and AI companies, has supported self-regulation of artificial intelligence. The companies have signed a voluntary agreement on safety checks, but it does not amount to a legally binding rule. Trump described the deal as “morally binding.”
Tech giants tasked with managing their own AI
The agreement was reached after Trump met with representatives from companies such as OpenAI, Google, Meta, Anthropic, xAI, and Nvidia at the White House.
The document, titled the White House Accord on Super Intelligence: A Joint Commitment on Frontier Responsibilities, sets out principles for companies to secure the safety of their most advanced systems through their own internal processes.
However, the accord is neither a law nor a regulation which would compel technology companies to adhere to new obligations. It remains voluntary.
After the meeting, Trump stated that, in his view, the tech sector itself should be largely responsible for controlling AI.
Agreement outlines four layers of oversight
The signed document proposes a system with several layers of safety mechanisms.
Companies are expected to focus on internal checks during the development and use of models. The next layer involves a dedicated internal team overseeing the operation of these safety mechanisms and addressing potential issues.
The accord also includes provisions for independent external audits. Ultimately, company board committees will review the results of these safety checks.
The document also addresses areas such as cybersecurity, biological and chemical risks, as well as scenarios where AI could gain unauthorized access to technical systems.
Legal framework remains undecided
The wording at the end of the agreement is significant. The document acknowledges that it may be appropriate in the future to enshrine these measures in laws or regulations.
Currently, however, the agreement does not impose any legal obligations.
Trump is therefore taking a different approach from the idea that the government should immediately establish new safety standards for advanced AI by law.
The issue of AI regulation is not new in the US. Even before this latest agreement, Trump had rejected sweeping new restrictions on AI development. Aktualizovane.sk has recently covered his stance. Trump rejects AI regulation and discusses new boundaries.
Trump introduces “super intelligence” terminology
The White House meeting also introduced a change in terminology for the US administration.
Trump signed an executive order directing the federal government to officially use the term “Super Intelligence” instead of the usual “artificial intelligence” or “AI” in its communications.
This represents a terminological shift within federal government, but the name change does not affect the technical features or risks associated with artificial intelligence systems.
Debate over AI safety continues within tech sector
The US administration’s approach comes amid increasing debate over how to control the development of the most advanced AI models.
Some tech industry figures warn of potential safety risks and call for stricter oversight mechanisms. Others emphasize the need for rapid technological advancement, expressing concern that excessive regulation could harm US competitiveness.
A similar issue is under discussion in Europe. Aktualizovane.sk has reported that Slovakia is preparing a state oversight system for artificial intelligence, in line with European regulations. Slovakia is developing its own AI oversight system.
The difference between the American and European approaches thus remains significant. While Washington’s latest agreement relies on voluntary commitments from tech companies, the European framework also involves public oversight and binding rules.
Whether self-regulation will be sufficient remains an open question. The development of advanced AI systems will also be closely linked to technological competition between the United States and China. The US and China are entering a new era of technological rivalry.
READ ALSO
EU softens AI rules: tech giants gain concessions
This article was translated from the original Slovak version with the assistance of AI.
